Browse Topic: Real-time data
This paper presents control and estimation approaches for multiple vehicles to cooperatively sample atmospheric variables, with a focus on wind estimation, in complex environments. The technology is encapsulated in the WINDSENSE system. The collected data could be used to initialize weather models for nowcasting or forecasting, assess the fidelity of a meteorological model, assist in understanding plume dynamics or tracking plumes, provide real-time data for fire controls or wildfire fighting, or locate the source of a chemical, biological, radiological, or nuclear (CBRN) source. The wind estimation approach utilizes a Bayesian formulation with a process model found using system identification techniques. The model structure of the identified dynamics builds on prior work by the authors and combines first-principles and experimental data collection to generate a model that is valid over a wide range of the flight envelope. This enables the wind estimator to also be viable over the modeled domain, allowing for wind estimation at trim and quasi-trim conditions and during dynamic maneuvers. Performance of the wind estimator is validated in field testing via comparison with a research-grade ultrasonic anemometer. A decentralized controller enables coordinated flight amongst multiple vehicles to maintain a desired formation geometry and reshape the formation to optimize data collection for different environments or mission objectives. Sampling strategies for a swarm of vehicles are developed through analysis of large-eddy simulation studies of wind and particle dispersion using the PALM facility.
As electric vehicles (EVs) become more advanced, so ensuring the reliability of critical components like the motor and Motor Control Unit (MCU) is essential. This paper presents a digital twin model designed to predict failures in motor and MCU components using machine learning. The approach focuses on detecting early signs of failure through real-world data and advanced analytics. We collected thermal and performance data from field vehicles, capturing both normal (healthy) and abnormal (faulty) operating conditions. Using this dataset, we developed and trained an Auto Encoder-based machine learning model that learns what “normal” looks like and flags deviations as potential issues. One key outcome of this study is the successful early prediction of Insulated Gate Bipolar Transistor (IGBT) degradation, where the system identified subtle behavioral changes long before any visible failure symptoms appeared. This digital twin acts as a virtual replica of the physical components, continuously monitoring and comparing real-time data to the learned normal behavior. It serves as a powerful tool for predictive maintenance, helping to reduce downtime, avoid unexpected failures, and optimize vehicle performance. The strength of this work lies in combining actual component-level data with a robust machine learning pipeline to create a scalable and practical failure prediction system. We are currently expanding this model to cover a wider range of failure scenarios for both motors and MCUs. This study offers a significant step toward smarter, more reliable electric vehicles by enabling early detection of potential failures through digital twins and AI.
In the context of increasing global energy demand and growing concerns about climate change, the integration of renewable energy sources with advanced modelling technologies has become essential for achieving sustainable and efficient energy systems. Solar energy, despite its considerable potential, continues to face challenges related to performance variability, limited real-time insights, and the need for reactive maintenance. To overcome these barriers, this work presents a Digital Twin framework aimed at optimizing solar-integrated energy systems through real-time monitoring, predictive analytics, and adaptive control. This work presents a Digital Twin framework designed to address the challenges of designing, operating, maintaining, and estimating renewable energy systems, specifically solar power, based on dynamic load demand. The framework enables real-time forecasting and prediction of energy outputs, ensuring systems operate efficiently and maintain peak performance across diverse conditions. The proposed methodology mirrors the physical system using real-time data inputs, environmental conditions, and physics-based models to create a high-fidelity virtual replica. This allows for dynamic analysis of energy flows, load forecasting, system performance prediction, and scenario testing to optimize design and operational strategies. By integrating predictive analytics, Digital Twin adapts to changing conditions, enabling proactive maintenance, fault detection, and system calibration to meet future load demands. Experimental validation demonstrates that the framework improves system efficiency, adaptability, and reliability, with scalable applications for both centralized and decentralized energy systems. Additionally, its integration with cloud-based platforms and IoT technologies enables real-time monitoring, facilitating continuous optimization and data-driven decision-making. This Digital Twin approach provides an intelligent, data-driven solution for the renewable energy sector, facilitating sustainable, resilient, and efficient energy infrastructures that can reliably meet evolving load demands while optimizing performance throughout their lifecycle.
Without reliability and signal integrity, aerospace communications risk severe signal degradation and reduced security, posing risks to both personnel and mission-critical data. These challenges are particularly critical for applications that depend on military aircraft, satellite communications, and unmanned aerial vehicles (UAVs). As global demand for real-time data continues to surge, communication infrastructure requires regular maintenance and upgrades to maintain secure and reliable performance.
Tillage, a fundamental agricultural practice involving soil preparation for planting, has traditionally relied on mechanical implements with limited real-time data collection or adjustment capabilities. The lack of real-time data and implement statistics results in fleet managers struggling to track performance, driver behavior, and operational efficiency of the implements. Lack of data on vehicle performance can result in unexpected breakdowns and higher maintenance costs, ensuring compliance with regulations is challenging without proper data tracking, potentially leading to fines and legal issues. Bluetooth-enabled mechanical implements for tillage operations represent an emerging frontier in precision agriculture, combining traditional soil preparation techniques with modern wireless technology. Implement mounted battery powered BLE (Bluetooth Low Energy) modules operated by solar panel based rechargeable batteries to power microcontroller. When Implement is operational turns module active and establishes communication with BLE capable wireless controller to share implement statistics and parameters. Sensors like magnetic pickup rotary shaft speed and accelerometer sensors interfaced with module to acquire implement working shaft speed, depth of implement operation in field. Based on dynamic data collected by module such as hours of usage, Trip hours, Oil change alert, maintenance alerts made available to the fleet managers. BLE module transmits acquired data to the tractor mounted wireless controller which makes data available on cloud, to fleet managers using dedicated applications to track all tillage implement statistics. Also, this solution helps implement manufacturers to track implement usage and avoid false warranty claim issues.
Most current AI models are based on static datasets, limiting their adaptability and real-time diagnostic potential. To address this gap, researchers have developed a novel proof-of-concept deep learning model that leverages real-time data to assist in diagnosing nystagmus — a condition characterized by involuntary, rhythmic eye movements often linked to vestibular or neurological disorders.
As mission-critical systems demand more processing power, real-time data movement, and multi-domain interoperability, rugged embedded systems are being transformed. Today's military and aerospace applications increasingly demand the merging of AI computing, enhanced sensor interfaces, and cybersecurity - all under harsh environmental conditions. At the heart of this evolution is the 3U OpenVPX form factor, a modular, compact, and ruggedized hardware standard and increasingly the SOSA aligned subset of the architecture. However, next-generation systems need to go further: supporting higher bandwidth, better thermal efficiency, improved security, while maintaining multi-vendor interoperability and long-term sustainability. We'll discuss some of today's enclosure solutions as well as emerging technologies.
This paper explores the integration of Microsoft Power BI into Model-Based Systems Engineering (MBSE) workflows, specifically within a Model-Based Product Line Engineering (MBPLE) context. Power BI provides a versatile platform for visualizing, analyzing, and manipulating data, enabling users to configure system variants outside traditional MBSE environments while maintaining integration back into the original MBSE model. This approach enhances collaboration between engineering and business disciplines, improves decision-making with real-time data analysis, and allows users to configure and evaluate multiple system variants efficiently. Additionally, the paper discusses how Power BI’s interactive dashboards facilitate better accessibility and analysis, bridging the gap between technical teams and non-technical stakeholders. Future work will focus on improving data pipeline automation and incorporating feature performance metrics to enable real-time trade study analysis, further enhancing system optimization and long-term decision-making.
This paper aims to explore the application of machine learning techniques to the analysis of road suspension systems, with particular emphasis on mechanical leaf spring suspensions. These systems are essential for vehicle performance, as they guarantee comfort and stability while driving, and they have an intrinsically complex and non-linear dynamic behavior. Because of this complexity, traditional approaches often prove costly and insufficient to represent operating conditions. In this context, machine learning techniques stand out for their ability to learn patterns from experimental data, allowing the modelling of non-linear phenomena that characterize road implement suspensions. One of the main contributions of this study is the demonstration that machine learning algorithms are capable of identifying complex patterns to represent the behavior of the system, as well as facilitating the detection of anomalies and potential faults in the suspension system, contributing to predictive maintenance. The results indicate that the application of machine learning algorithms not only improves the accuracy of suspension performance analysis, but also offers an innovative approach to diagnosing and identifying problems. With the ability to process and analyze data in real-time, these technologies can be integrated into vehicle monitoring systems, allowing for quick and effective interventions. In conclusion, the use of machine learning in the analysis and design of road suspension systems represents a significant advance in automotive engineering. The research highlights the emergence of new research horizons in this area, suggesting that the combination of engineering knowledge and artificial intelligence can open up new frontiers for the development of more efficient and safer suspension systems.
For mature virtual development, enlarging coverage of performances and driving conditions comparable with physical prototype is important. The subjective evaluation on various driving conditions to find abnormal or nonlinear phenomena as well as objective evaluation becomes indispensable even in virtual development stage. From the previous research, the road noise had been successfully predicted and replayed from the synthesis of system models. In this study, model based NVH simulator dedicated to virtual development have been implemented. At first, in addition to road noise, motor noise was predicted from experimental models such as blocked force and transfer function of motor, mount and body according to various vehicle conditions such as speed and torque. Next, to convert driver’s inputs such as acceleration and brake pedal, mode selection button and steering wheel to vehicle’s driving conditions, 1-D performance model was generated and calibrated. Finally, the audio and visual feedback correspondent with driver’s input was represented in the simulator with real-time data network between various hardware and software. To validate the simulator, subjective evaluation was performed with so-called virtual vehicles by changing tires, rubber mounts, suspension and body on various roads, speed and torque, which showed contextual results with physical prototypes. In conclusion, the NVH simulator equipped with consistent experimental and simulation models could be utilized to find and improve abnormal or nonlinear phenomena in virtual vehicle development stage, which can help to frontload vehicle development.
U.S. Army Combat Capabilities Development Command Chemical Biological Center (DEVCOM CBC) researchers are developing a way to scan for chemical biological agent on surfaces on the fly. Literally on the fly as it consists of an AI-enabled spectrometer mounted on an unmanned aerial vehicle (UAV) or unmanned ground vehicle (UGV) sending back vital data in real time. It is called Hyperspectral Threat Anomaly Detection, or HyperThreAD for short.
The deployment of PEM fuel cell systems is becoming an increasingly pivotal aspect of the electrification of the transport sector, particularly in the context of heavy-duty vehicles. One of the principal constraints to market penetration is durability of the fuel cell which hardly meets the expected targets set by the vehicle manufacturers and regulatory bodies. Over the years, researchers and companies have faced the challenge of developing reliable diagnostic and condition monitoring tools to prevent early degradation and efficiency losses of fuel cell stack. The diagnostic tools for fuel cell rely usually on model-based, data driven and hybrid approaches. Most of these are mainly developed for stationary and offline applications, with a lack of suitable methods for real-time and vehicle applications. The work presented is divided into two parts: the first part explores the main degradation conditions for a PEMFC and characteristics, advantages, and application limits of the main methodologies for fuel cell diagnostic, while in the second part the features and the development process of an innovative, real-time, and on-board health and condition monitoring system, based on electrochemical impedance spectroscopy (EIS), are presented. The new innovative tool allows to detect, identify and isolate degradation, faults and non-optimal conditions. The computational performance and reliability of the diagnostic tool are tested and validated through experimental tests carried out in the laboratory on single cell and short fuel cell stack over a wide range of operating conditions and under specific sub-optimal/fault states such as drying, flooding and reactants starvation. The condition and health of PEMFC are estimated using specific health indicators for the most common root causes of faults such as drying, flooding, catalyst poisoning and anode/cathode starvation.
To meet the requirements of high-precision and stable positioning for autonomous driving vehicles in complex urban environments, this paper designs and develops a multi-sensor fusion intelligent driving hardware and software system based on BDS, IMU, and LiDAR. This system aims to fill the current gap in hardware platform construction and practical verification within multi-sensor fusion technology. Although multi-sensor fusion positioning algorithms have made significant progress in recent years, their application and validation on real hardware platforms remain limited. To address this issue, the system integrates BDS dual antennas, IMU, and LiDAR sensors, enhancing signal reception stability through an optimized layout design and improving hardware structure to accommodate real-time data acquisition and processing in complex environments. The system’s software design is based on factor graph optimization algorithms, which use the global positioning data provided by BDS to constrain the drift of IMU and LiDAR data, ensuring that the system can maintain accurate positioning through IMU and LiDAR collaboration, even when GNSS signals are limited or completely unavailable. Experimental results show that the system’s 3D positioning error in shaded environments is controlled within 7 cm, with a convergence time of no more than 40 seconds. Further statistical analysis reveals a root mean square error (RMSE) of approximately 8 cm and a standard deviation (STD) of 2 cm. During the simulated indoor-outdoor scene transition test, the system’s relative pose error remains stable within 10 cm, demonstrating its adaptability and robustness in diverse and complex scenarios. This study provides a technical reference for the hardware construction and system validation of multi-sensor fusion technology on autonomous driving platforms.
Intelligent Structural Health Monitoring (SHM) of bridge is a technology that utilizes advanced sensor technology along with professional bridge engineering knowledge, coupled with machine vision and other intelligent methods for continuously monitoring and evaluating the status of bridge structures. One application of SHM technology for bridges by way of machine learning is in the use of damage detection and quantification. In this way, changes in bridge conditions can be analyzed efficiently and accurately, ensuring stable operational performance throughout the lifecycle of the bridge. However, in the field of damage detection, although machine vision can effectively identify and quantify existing damages, it still lacks accuracy for predicting future damage trends based on real-time data. Such shortfall l may lead to late addressing of potential safety hazards, causing accelerated damage development and threatening structural safety. To tackle this problem, this study designs a deep learning model based on temporal information to solve the problem of predictive damage development, achieving early warning and dynamic evaluation effects. This study focuses on concrete crack development, and the CrackAE model is based on traditional semantic segmentation models and conditional autoencoder architecture. The model consists of an encoder and a decoder. The encoder accepts image data and outputs a feature map. The future map along with the conditional vector encoded based on physical temporal information, serves as the input to the decoder. The output of decoder is the development state of the crack at the specified prediction time. The model achieved an accuracy of 94.6% in real bending failure tests of concrete beams, indicating that the model meets high-precision prediction requirements. This validates the feasibility of deep learning in predicting damage development and provides new ideas for data collection and prediction in actual bridge maintenance.
Items per page:
50
1 – 50 of 289